Abstract
Empirical studies of perception must use the logic of everyday non-technical conceptions of perception as their unquestioned background. This is because the phenomena to be studied are defined and individuated on the basis of such basic understanding. Thus the methods of neurobiology exclude reductionist accounts from the outset, implicitly if not explicitly. It is further argued that the concepts of neural and mental representation, while not confused per se, presuppose a general picture where perception as a whole is viewed in the light of teleology. References are made to discussions by Bennett and Hacker, Paul Churchland, and Peter Winch